Enabling 3D Ultrasound Procedure Guidance through Enhanced Visualization

Brattain, Laura J.; Vasilyev, Nikolay V.; Howe, Robert D.

doi:10.1007/978-3-642-30618-1_12

Cited by 7 publications

(8 citation statements)

References 19 publications

(21 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this study, we designed and evaluated a CNN‐based method for segmenting needle‐like tools in 2D US images in near real‐time (approximately 50 ms per tool), demonstrated on unseen data from phantoms and five different interventional applications from four anatomical sites. The widespread use of real‐time 2D US imaging 1,2 makes the algorithm clinically relevant for a large variety of minimally invasive percutaneous interventions that require accurate placement of tools to achieve desired diagnostic and therapeutic results. The near real‐time segmentation of the algorithm provides the potential for live localization and better visualization of tools during insertion for interventional procedures.…”

Section: Discussionmentioning

confidence: 99%

“…Both prostate and gynecologic brachytherapy images were acquired using a BK ProFocus 2202 US system with an 8848 endocavity probe (BK Medical, MA, USA). Pixel sizes for the brachytherapy images ranged between 0.126 × 0.126 mm 2 and 0.212 × 0.212 mm 2 . Both brachytherapy procedures had needles inserted nearly parallel to the linear US imaging face, often resulting in reverberation artifacts.…”

Section: B Datasetsmentioning

confidence: 99%

“…Although three-dimensional (3D) US and real-time 3D US imaging methods have emerged, adoption of these technologies for minimally invasive interventions has been limited and real-time two-dimensional (2D) US remains the predominant modality for guidance of needles and therapy applicators into the body. 1,2 Despite the benefits of minimally invasive percutaneous approaches, one factor that has limited the guidance accuracy of these techniques is the ability to localize needle-like interventional tools, such as needles and therapy applicators, quickly and accurately in the standard 2D US images while in the intraoperative environment. 3,4 Diagnostic and therapeutic cancer procedures where accurate localization is essential include brachytherapy, solid organ ablation, and biopsy.…”

Section: Introductionmentioning

confidence: 99%

“…Image-based approaches for tool guidance, avoiding the requirement for additional equipment, signal processing, and set-up, have been proposed for general tool segmentation in US images, leveraging techniques based on image properties, including projections, [8][9][10][11] random sample consensus (RANSAC), [12][13][14] filtering, 13,[15][16][17] and Hough or Radon transforms, [18][19][20][21][22] or physical properties, such as analyses of motion, [22][23][24][25][26] beam steering, 27 and circular wave generation. 28 Many of these algorithms were developed for 3D US images; [8][9][10][11][12][13][14][15]20,21,23 however, real-time 2D US is the clinical standard for guiding tools during image-guided minimally invasive interventions at most institutions 1,2 and therefore is the focus of our work presented in this study. Many of the published approaches were tested only on phantom or ex vivo tissue images, [8][9][10][11][12][13][14]…”

Section: Introductionmentioning

confidence: 99%

“…This is particularly advantageous for guidance during minimally invasive percutaneous interventional techniques, which offer reduced recovery times and complications relative to open surgery. Although three‐dimensional (3D) US and real‐time 3D US imaging methods have emerged, adoption of these technologies for minimally invasive interventions has been limited and real‐time two‐dimensional (2D) US remains the predominant modality for guidance of needles and therapy applicators into the body 1,2 . Despite the benefits of minimally invasive percutaneous approaches, one factor that has limited the guidance accuracy of these techniques is the ability to localize needle‐like interventional tools, such as needles and therapy applicators, quickly and accurately in the standard 2D US images while in the intraoperative environment 3,4 .…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Deep learning segmentation of general interventional tools in two‐dimensional ultrasound images

et al. 2020

View full text Add to dashboard Cite

Purpose: Many interventional procedures require the precise placement of needles or therapy applicators (tools) to correctly achieve planned targets for optimal diagnosis or treatment of cancer, typically leveraging the temporal resolution of ultrasound (US) to provide real-time feedback. Identifying tools in two-dimensional (2D) images can often be time-consuming with the precise position difficult to distinguish. We have developed and implemented a deep learning method to segment tools in 2D US images in near real-time for multiple anatomical sites, despite the widely varying appearances across interventional applications. Methods: A U-Net architecture with a Dice similarity coefficient (DSC) loss function was used to perform segmentation on input images resized to 256 × 256 pixels. The U-Net was modified by adding 50% dropouts and the use of transpose convolutions in the decoder section of the network. The proposed approach was trained with 917 images and manual segmentations from prostate/gynecologic brachytherapy, liver ablation, and kidney biopsy/ablation procedures, as well as phantom experiments. Real-time data augmentation was applied to improve generalizability and doubled the dataset for each epoch. Postprocessing to identify the tool tip and trajectory was performed using two different approaches, comparing the largest island with a linear fit to random sample consensus (RAN-SAC) fitting. Results: Comparing predictions from 315 unseen test images to manual segmentations, the overall median [first quartile, third quartile] tip error, angular error, and DSC were 3.5 [1.3, 13.5] mm, 0.8 [0.3, 1.7]°, and 73.3 [56.2, 82.3]%, respectively, following RANSAC postprocessing. The predictions with the lowest median tip and angular errors were observed in the gynecologic images (median tip error: 0.3 mm; median angular error: 0.4°) with the highest errors in the kidney images (median tip error: 10.1 mm; median angular error: 2.9°). The performance on the kidney images was likely due to a reduction in acoustic signal associated with oblique insertions relative to the US probe and the increased number of anatomical interfaces with similar echogenicity. Unprocessed segmentations were performed with a mean time of approximately 50 ms per image. Conclusions: We have demonstrated that our proposed approach can accurately segment tools in 2D US images from multiple anatomical locations and a variety of clinical interventional procedures in near real-time, providing the potential to improve image guidance during a broad range of diagnostic and therapeutic cancer interventions.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: B Datasetsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Deep learning segmentation of general interventional tools in two‐dimensional ultrasound images

et al. 2020

View full text Add to dashboard Cite

show abstract

Instrument Tracking and Visualization for Ultrasound Catheter Guided Procedures

Brattain

Loschak²,

Tschabrunn

et al. 2014

Augmented Environments for Computer-Assisted Interventions

Self Cite

View full text Add to dashboard Cite

We present an instrument tracking and visualization system for intra-cardiac ultrasound catheter guided procedures, enabled through the robotic control of ultrasound catheters. Our system allows for rapid acquisition of 2D ultrasound images and accurate reconstruction and visualization of a 3D volume. The reconstructed volume addresses the limited field of view, an inherent problem of ultrasound imaging, and serves as a navigation map for procedure guidance. Our robotic system can track a moving instrument by continuously adjusting the imaging plane and visualizing the instrument tip. The overall instrument tracking accuracy is 2.2mm RMS in position and 0.8° in angle.

show abstract

Spatial calibration of a 2D/3D ultrasound using a tracked needle

Vasconcelos¹,

Peebles

Ourselin³

et al. 2016

Int J CARS

View full text Add to dashboard Cite

PurposeSpatial calibration between a 2D/3D ultrasound and a pose tracking system requires a complex and time-consuming procedure. Simplifying this procedure without compromising the calibration accuracy is still a challenging problem.MethodWe propose a new calibration method for both 2D and 3D ultrasound probes that involves scanning an arbitrary region of a tracked needle in different poses. This approach is easier to perform than most alternative methods that require a precise alignment between US scans and a calibration phantom.ResultsOur calibration method provides an average accuracy of 2.49 mm for a 2D US probe with 107 mm scanning depth, and an average accuracy of 2.39 mm for a 3D US with 107 mm scanning depth.ConclusionOur method proposes a unified calibration framework for 2D and 3D probes using the same phantom object, work-flow, and algorithm. Our method significantly improves the accuracy of needle-based methods for 2D US probes as well as extends its use for 3D US probes.

show abstract

Enabling 3D Ultrasound Procedure Guidance through Enhanced Visualization

Cited by 7 publications

References 19 publications

Deep learning segmentation of general interventional tools in two‐dimensional ultrasound images

Deep learning segmentation of general interventional tools in two‐dimensional ultrasound images

Instrument Tracking and Visualization for Ultrasound Catheter Guided Procedures

Spatial calibration of a 2D/3D ultrasound using a tracked needle

Contact Info

Product

Resources

About